• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

LM3886 - Noisy signal unless inputs shorted

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Okay, me again. Near-absolute beginner. Knowledge limited to mostly soldering and basic DMM / scope usage.

I now have two Chipamp.com LM3886 boards with stock resistors/gain and two I've modified with Linkwitz' values
after discussion here in another thread.

My question is: On both sets of boards, I get a lot of (noise?) on the scope unless the input leads are shorted. This seems to happen whether or not they are attached to an audio source. Is this normal?

Here is one board with Linkwitz values of resistors swapped in.
(I haven't yet swapped Rm, the Mute resistor with the Linkwitz value one yet). Input leads shorted.
View attachment DS0000.BMP


Same board, (Linkwitz mods) with input leads not shorted or connected to anything.
View attachment DS0001.BMP


Board with stock components. Input leads shorted.
View attachment DS0002.BMP


Board with stock components. Input leads not shorted or connected to anything.
View attachment DS0003.BMP

I forgot to put the scope into the measurement screen when taking the first
screenshots. The PtoP voltage on the flatter signal on that board is about 2.2 to 2.4 mV .

Nothing is connected to earth/ground. I wasn't sure how to do that for basic testing.
Without meaning to sound ungrateful, last time I asked about basic grounding connections, the discussion
devolved into a debate about the perfect method, and I never got an answer I could understand at
my level, so I just let it go.

In that earlier thread, coaxial cable was recommended for input leads, but using it just seemed to
make things worse.


Thanks.
 
Last edited:
No volume pot.

I can't remember the measurements. I tried with a Sony CD player directly connected. I could put it back on the scope if you can suggest what might be a good test. I have an audio test CD here, or I could download and burn something, if really necessary.


Do you have a volume pot on the input? How much noise are you getting with it connected to a source?
 
Well, I was using a little test jig that a friend had created for me. RCA inputs on one end, and terminals on the other to connect to the amp inputs. This jig has a POT volume control, and it worked fine. Later, it seemed like there was a lot of noise on the scope, even with the jig (and its POT). Can you suggest some tests to run?

Would it be better to use a smartphone or something as a source? They have lower output voltage, don't they?


Do you have it connected to a preamp? You need to attenuate the input, the noise on your scope is whatever is being picked up on the input amplified to the maximum.
 
Last edited:
You have proved that the amp is not making the noise by shorting the inputs. You really need to attach a volume control permanently on the input before you go any further and definitely before you connect a speaker. A 10K log pot would be ideal. Also before you connect a speaker check for any DC offset on the output
 
Which signal do I use to measure DC offset? I'll do that and post the pics back here. Oh and..um...does it matter for what I just posted above whether the scope is set to DC or AC coupling mode?


You have proved that the amp is not making the noise by shorting the inputs. You really need to attach a volume control permanently on the input before you go any further and definitely before you connect a speaker. A 10K log pot would be ideal. Also before you connect a speaker check for any DC offset on the output
 
When you say "no input", do you mean with input wires shorted?

If so, you should be able to make out those numbers from the above screenshots.
It fluctuates between About 2.8 and 3.8 mV peak to peak. For both boards I tested. I think the scope was in DC coupling mode for that measurement. I believe that is the icon being displayed in the above screenshots.

This is confusing. Definitions on the web for "DC offset" displaying on a scope are different than what you're describing.

I tried attaching the jig with the POT on it. This time, I got less a noisy waveform, but I'm guessing still messed up. It seems like the input wires are extremely sensitive to attachment quality and/or being moved. Is this normal? The POT on the jig has the following markings on it:

1 Borton 2 CW 3
3006P 103
N036 (upside down)

Does this mean anything to you?

DC offset is the voltage measured across the speaker terminals with no input. Ideally it will be zero, but is more likely to be around +/- 10mV to +/- 50mV. Use a multimeter if you have one or your scope set to dc coupling, for your noise measurements ac coupling is fine
 
Last edited:
Oh, it looks like the POT on that jig I have is a Bourns 3006P 103. It's a 50K 3/4W model. The weird thing about it is that the scope signal doesn't seem to change no matter how much I adjust the POT screw. This is strange, because a while back, when I played music through either amp board, the POT did adjust the volume up and down.
 
Don´t overthink it, you are not testing that amp the way you will actually use it.

Just connect a speaker at its output, a sound source (PC/MP3 player/Smartphone/CD player) to a volume pot, between 10k and 50k, Audio/Log taper preferred, and from it to the amp, using a coaxial cable.
Also coaxial cable and properly soldered connectors between sound source and potentiometer (a real one, with metallic case soldered to ground)
Also place the amplifier board inside a metallic case or at least on a piece of wood, to which you glue a piece of kitchen type aluminum paper, then ground it to amp ground with a piece of wire, and place an insulator (cardboard or a magazione will do) on aluminum so it does not short PCB bottom.
Play some music and *then* worry or not about Hum and external noise.
 
Don´t overthink it, you are not testing that amp the way you will actually use it.

Just connect a speaker at its output, a sound source (PC/MP3 player/Smartphone/CD player) to a volume pot, between 10k and 50k, Audio/Log taper preferred, and from it to the amp, using a coaxial cable.

Like I said, I used coaxial cable previously, and it just seemed to create more noise. When I removed it, I got less noise.

Also coaxial cable and properly soldered connectors between sound source and potentiometer (a real one, with metallic case soldered to ground)
Also place the amplifier board inside a metallic case or at least on a piece of wood, to which you glue a piece of kitchen type aluminum paper, then ground it to amp ground with a piece of wire, and place an insulator (cardboard or a magazione will do) on aluminum so it does not short PCB bottom.
Play some music and *then* worry or not about Hum and external noise.

I'll try that. I hope finding a volume POT at a bricks and mortar store doesn't take too long.
 
Is there any reason I should not use the one Chipamp.com sold? It was an Alps Blue Velvet RK27. When I look at Mouser, they have about 10 different version of it. Which one should I buy? I'd prefer a through hole model if it has larger (easier to solder/desolder) terminals.

Does anyone still have any of the matching PCBs that Chipamp.com made for those POTs.
 
scottjoplin:

Thanks. Unfortunately, they are on backorder at Mouser in Canada and the US. Digikey doesn't sell them either. Where else can I look in North America that might have them?

As for the PCBs, will a cheap Chinese one work, or should I look for something else? I don't suppose anyone knows where I could find one of Chipamp's own boards? Are the ALPS pots they sell on eBay from Asia counterfeit or are they genuine?
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.